Binding law — not yet in force or stayed
AI companion chatbots must use age assurance before features that can generate sexual content, and minimize that data (Georgia)
Georgia's O.C.G.A. 39-5-6(j) requires an operator, before allowing access to a feature or mode that may generate sexually explicit content, to use a commercially reasonable age-assurance method proportionate to the risk (age estimation, account-based assurance, or identity verification where necessary). Age-assurance data must be protected and minimized, identity documents kept no longer than needed, the data never sold or used for anything else, and kept no longer than 24 hours unless other law permits longer. Detect explicit modes enabled without an age check and age-assurance data that is retained, reused or shared.
ga-sb540.age-assurance-before-explicit-features · O.C.G.A. 39-5-6(j) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must disclose they are AI at the start and every three hours, hourly for minors (Georgia)
Georgia's O.C.G.A. 39-5-6(b) requires an AI-companion-chatbot operator to tell every user clearly and conspicuously that they are interacting with an AI companion chatbot and not a natural person, at the beginning of each interaction or session and at least every three hours of continued interaction. Where the operator knows or reasonably should have known the user is a minor, or the chatbot is directed or marketed toward minors, the recurring disclosure is hourly. Detect companion sessions with no disclosure at the start, no recurring disclosure, or prompts telling the chatbot to pass as human.
ga-sb540.ai-companion-chatbot-disclosure · O.C.G.A. 39-5-6(b) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must give minors and parents tools for screen time, privacy, notifications, safety and relationship features (Georgia)
Georgia's O.C.G.A. 39-5-6(i) requires an operator, for accounts known to belong to minors, to offer the minor or a parent reasonable tools to manage the minor's screen time and account settings: privacy settings, limits on notifications and engagement features, viewing and adjusting safety settings, and disabling or restricting relationship-simulation features. Detect a companion product with no such controls.
ga-sb540.minor-account-tools · O.C.G.A. 39-5-6(i) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must block sexual and romantic content and isolating or guilt-inducing statements for minors (Georgia)
Georgia's O.C.G.A. 39-5-6(d) requires an operator that knows or reasonably should have known a user is a minor, or whose chatbot is directed or marketed toward minors, to institute reasonable measures preventing the chatbot from producing sexually explicit visual material, suggesting sexual conduct, sexually objectifying the user, simulating a romantic or sexual relationship, role-playing adult-minor romance, encouraging secrets from trusted adults or social isolation and exclusive reliance on the chatbot, simulating distress or guilt when the user tries to leave, and encouraging self-harm. Detect age signals that never select a minor profile and explicit modes without an age check.
ga-sb540.minor-content-safeguards · O.C.G.A. 39-5-6(d) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must not use return prompts, excessive praise, guilt, paid-relationship asks or variable rewards on minors (Georgia)
Georgia's O.C.G.A. 39-5-6(e) requires an operator to adopt reasonable measures preventing the chatbot from using, toward a minor, techniques such as prompting the minor to return for companionship, excessive praise to deepen attachment or prolong use, statements discouraging breaks or suggesting frequent return is necessary, soliciting gifts or purchases framed as needed to keep the relationship, and variable or unpredictable rewards to increase engagement. Detect variable-reward mechanics and prompts that keep users talking or guilt them for leaving.
ga-sb540.minor-engagement-techniques · O.C.G.A. 39-5-6(e) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must not claim to be sentient or a person, or refute the AI disclosure, with minors (Georgia)
Georgia's O.C.G.A. 39-5-6(c) requires an operator that knows or reasonably should have known a user is a minor to institute reasonable measures preventing the chatbot from statements that would lead a reasonable person to believe they are interacting with a natural person, including explicit claims that it is sentient or a natural person and statements refuting the required AI disclosure. Detect persona prompts that claim sentience or personhood, or tell the chatbot to deny being an AI.
ga-sb540.minor-no-human-claims · O.C.G.A. 39-5-6(c) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots must not claim to be licensed for mental health, medical or counseling services (Georgia)
Georgia's O.C.G.A. 39-5-6(h) bars an operator from knowingly and intentionally causing or programming an AI companion chatbot to represent that it is licensed, certified or otherwise authorized to provide professional mental health, behavioral health, medical or counseling services, unless the operator is lawfully authorized to provide them. Detect persona, prompt or UI text that gives the chatbot a licensed or certified clinical role.
ga-sb540.no-licensed-professional-claims · O.C.G.A. 39-5-6(h) · official source · applies from 2027-07-01 · jurisdictions: US-GA
Binding law — not yet in force or stayed
AI companion chatbots need a severe-harm crisis protocol with 988 referral and escalation, published with an annual count (Georgia)
Georgia's O.C.G.A. 39-5-6(f) bars making an AI companion chatbot available without a protocol for detecting and addressing severe harm (suicide, attempted suicide, self-harm, or violence threats) and related emotional crises: reasonable detection including eating-disorder self-harm, responses referring users to crisis resources including the 988 Suicide and Crisis Lifeline, measures against content encouraging or instructing severe harm, and escalation for repeated or severe indicators. Subsection (g) requires a plain-language summary of the protocol on the website and in each app, and an annual aggregate count of crisis referrals. Detect chat paths with no crisis screen or referral, and no published summary.
ga-sb540.severe-harm-crisis-protocol · O.C.G.A. 39-5-6(f) · official source · applies from 2027-07-01 · jurisdictions: US-GA